How much red or near-infrared light are you actually getting?

Start with four quantities—not one headline number

PBM papers and product pages often use dose, fluence or energy density for a J/cm² value. Terminology is not always consistent. PLT uses incident radiant exposure here because it describes what the arithmetic establishes: energy arriving per unit area at a specified surface. It does not say how much energy is absorbed by tissue or whether a biological effect follows.

The basic calculation

If irradiance is reported in mW/cm², divide by 1,000 first because 1,000 milliwatts equals one watt.

Equal arithmetic does not guarantee equal biology

25 mW/cm² for 120 seconds and 100 mW/cm² for 30 seconds both calculate to 3 J/cm² incident radiant exposure. PBM reviews nevertheless describe dependence on irradiance, duration and other parameters—not just their product.[5] Use the equation to check arithmetic, not to declare two protocols therapeutically interchangeable.

Why distance changes the picture

An irradiance figure without distance is incomplete for a non-contact device. As you move away from an individual LED, its beam spreads and its contribution at a point generally changes. In an array, beams also overlap. The treatment-plane pattern therefore depends on LED spacing, optics, angle and distance—not just total electrical power.

A simple inverse-square calculation is not a safe substitute for a measured map of a nearby multi-LED panel. The array is not one ideal point source, and a consumer may be using it in the region where many diverging beams overlap. Moving farther away may reduce local peaks while changing the size and uniformity of the illuminated field. Only distance-specific measurements across the relevant plane can describe that field.

Masks illustrate the same issue in a different geometry. Close emitters remove the user-selected panel distance, but the face is curved and emitter-to-skin gaps can vary. The panel-versus-mask guide owns that form-factor decision; this page explains why neither geometry can be reduced to one output number.

Why one irradiance number can mislead

A number may be correct for the exact place and setup where it was measured while still being a poor description of the whole session. Hadis and colleagues found that 73% of the 74 PBM papers they reviewed did not report how light was measured; beam area, irradiance and pulse frequency were also frequently omitted.[3] Their review and earlier reporting guidance emphasise measurement at the target, beam area, beam profile, application technique and pulse parameters.[4]

The instrument is part of the result

“Measured” is not a complete method. A useful report identifies the instrument model, calibration status, spectral range or response, input optics, distance, angle, aperture and sampling approach. A spectrally resolved instrument can show how output is distributed by wavelength, while a broadband meter reduces its response to one reading. Neither label alone guarantees accuracy: detector response must suit the source, calibration matters, and a “spectrometer” is not necessarily a calibrated spectroradiometer.

That does not make manufacturer measurements inherently dishonest or useless. It means two readings made with different instruments or procedures should not be treated as if only the product changed. Manufacturer data can still be valuable when its method and conditions are explicit.

Continuous and pulsed output

For stable continuous output, the basic equation uses the irradiance operating during the full exposure. For pulsed output, ask whether the figure is peak irradiance during a pulse or time-averaged irradiance across the pulse-and-pause cycle. Peak irradiance multiplied by total session time can overstate incident radiant exposure if off-time is ignored. Pulse duration, repetition frequency, duty cycle and measurement method are therefore necessary context. Pulsing is not automatically a quality advantage, and evidence from one pulse pattern cannot be assigned to every device with a “pulse” button.

Real use adds another layer. Selected wavelength mode, dimming, warm-up or operating temperature, power source or battery state, angle, movement and actual distance may differ from the manufacturer’s measurement setup. Those factors do not prove a published value is wrong; they explain why a specification is a reference condition rather than a measurement of each home session. Research reporting guidance recommends checking output before and after exposures because stability itself should be verified.[4]

A real example from PLT’s panel data

BlockBlueLight’s two different values at the same stated distance make the method problem unusually visible.[9] They do not prove which number is the product’s “true irradiance” for every purpose. Infraredi’s larger headline cannot be compared cleanly because the checked table does not state distance or position.[8] PLT therefore does not turn these three values into a league table or use them to calculate recommended session times.

See how that uncertainty affects the Infraredi versus BlockBlueLight comparison and why irradiance remains outside the Australian panel shortlist table.

From incident exposure to a research protocol

A study protocol is more than a J/cm² value. To judge whether it resembles a consumer setup, look for wavelength and spectral width, source type, output mode, irradiance at the tissue, exposed area, distance or contact technique, time per exposure, number and spacing of sessions, anatomical site, population and measured outcome. Reporting guidance exists precisely because omissions prevent replication.[4]

Even complete incident-exposure reporting does not calculate energy at a deeper target. Biological tissue both absorbs and scatters light, and those optical properties vary with wavelength, tissue composition and structure.[7] Surface irradiance is therefore not the same quantity as absorbed energy in skin or irradiance at muscle, joint, brain or another deeper location. “Near-infrared penetrates deeper” is not a conversion factor.

PBM dose-response literature also cautions against treating more incident energy as automatically better. A review described biphasic and dose-rate-dependent responses, but its evidence at the time was largely laboratory and animal research and it did not establish one universal treatment window.[6] The consumer lesson is modest: do not infer an optimum, a danger threshold or a personal protocol from one generic chart or from the largest advertised irradiance.

What to look for before comparing two products

1. At what distance? Is it the distance the manufacturer actually tells you to use?
2. Where was the reading taken? Centre, peak, area average or a disclosed grid of points?
3. Across what area? Are panel dimensions being mistaken for illuminated treatment coverage?
4. With what instrument? Is the model, calibration and spectral suitability stated?
5. In which operating mode? Red only, near-infrared only, both, dimmed, continuous or pulsed?
6. Peak or average? For pulsed output, does the number include off-time?
7. Does the arithmetic use seconds and watts? Convert mW to W before calculating J/cm².
8. Is a study being borrowed? Does its device, geometry, protocol, anatomical site and outcome materially match?
9. Are the two numbers genuinely like-for-like? If methods differ or details are missing, leave the comparison unresolved.